# Visual estimation of amplitude

#### stefanS

##### Registered User
I have read that you can estimate the balance amplitude by visually looking at the balance arm position as it swings. Can anyone point to a diagram for a pocketwatch. I have only found so far one for a triple arm glucydur balance on wrist watch. I thought about taking video of the balance and slowing it down to check the amplitude.

Stefan

#### Smudgy

##### Registered User
I can't recall offhand the page numbers, but beleive Gazelly, Freid, and Daniels all cover the topic. The two armed balance is easier to determine rotation than the three armed. On the two armed balance the first crossing is at right angles to the at rest position and represents 1/2 turn. The second crossing looks like the same position as the at rest position and represents 1 turn. If the balance reaches the third crossing it will look similar to the first crossing and represents 1-1/2 turns.

#### HenryB

##### Registered User
I heard a new one today by John Cote, concerning a recent find.

I asked him how it runs, he replied,
it breathes
.

I know exactly what he means

#### stefanS

##### Registered User
With the movement running should the positions of the arm observed (due to persistence of vision) be the same i.e. visually like a fast blinking of the arm at the same angular position? Mine seems to have two very different positions. One at +90deg of rest and one at +105 or +110 deg. Watch keeps about +6 sec a day.

#### Smudgy

##### Registered User
It sounds like you have less balance action than desired. The balance should go through the 90 degree (half turn), then 180 degree (full turn) positions and continue further to the 90 degree position again (one and a half turns) if the watch is running well. Some watches, especially 7 jewel models, are difficult to get to the final one and a half turn positioning (the second 90 degree), but all need to get past the full turn position (180 degree) to be able to regulate the watch. It has to do with the amount of arc used during the escapement contact with the balance. A shorter action than a full turn (which is what it sounds like you have, maybe 3/4 turn) will usually result in a fast running watch that runs very differently between the short arcs and long arcs. You must be having a power loss somewhere in the watch, probably due to friction but it can also be from loose or small roller jewel oversized pivot holes and a large amount of other problems that display symptoms of low balance amplitude. Check that you have enough power transmitted to the escapewheel by checking for backward rotation when the train is spun and the mainspring click is off (the escapewheel should reverse direction at the end of the spinning in both directions). If not the loss will be in the train. If the escapewheel does the reversing at the end then the loss will be in the escapement or balance.

#### stefanS

##### Registered User
I was pointed to e-timer. I built a piezo sensor and amp and placed the watch in a watch box with padding so I can position the watch. I recorded the watch for a min or so in positions.

After setting regulator and regulating back and forth a few times I arrived at this for PU
+2.5 s/D
amplitude 324 deg
beat error 0.98 ms

DD and DU were close at ~-10s/D before adjusting
PU and PD were also close and near -10s/D

Noticed slight correlation between beat error and rate. Could not easily set both. I got better rates with higher ~1.5 ms errors or better beat error 0.7 ms with higher/lower rates. Above was a compromise. The regulator was very touchy. Is it supposed to be? Anyway for 110 yr old watch this is great to me. I know the beat error is a little high but is acceptable, Yes?

#### Smudgy

##### Registered User
The regulator shouldn't be touchy. The most likely cause fr it being touchy is that the hairspring needs to be trued. The hairspring should stay in the same position regardless of the position of the regulator. If it moves, even slightly, the hairspring needs to be adjusted. The hairspring out of true will usually cause a fast rate, and can cause a poor amplitude if it is badly out of true. It may also cause the hairspring to contact other things depending on the position of the regulator if it is badly out. I would suggest checking the hairspring for truth. Start by removing the hairspring from the balance and placing it in the balance cock to be sure that the collet centers over the balance jewel. Check that this condition is true for all positions of the regulator. Adjust the hairspring as needed. Then put the hairspring back on the balance and check that the balance is held vertically when installed on the cock. If it isn't held vertically then the spring needs centered. The spring should also be held flat and even over the balance. re-install and check the running condition.

#### Smudgy

##### Registered User
Newcomers are welcome.

RULES & GUIDELINES